How to drop from 6V to 5V a "clean" way?

[flotulopex]

I have made a kind of timer - more or less something like the ones for the chess players but for up to 5 players. My mother just drives me crazy when we play Rami and she wins but she takes almost triple time for reflection....

So to the project, there's a PIC, an LCD display, 5 LEDs, a piezo and a few external pull-up resistors.

On my breadboard, I had no problem to get the timer work properly. I have a lab power supply so no trouble here.

When I made the prototype, my thought went first to the power supply. Due to the boxe size contraints, I choose to go with four UM4-AAA batteries (I didn't want to go with accus and a charger).

After some tests, I had to modify quite a few things such as:
- external 10k pull-ups (the internal ones look to be to weak because both buttons just switch from time to time even if nobody presses them)
- addition of 100nF cap between Vdd and Vss (I never needed before) and another one between LCD and Vss (same)

Actually, the circuit consumes around 50mA when the display is ON and one LED is activated (normal operation). Strangely, when I start the timer, the current drops to 20mA. I'm still invastigating this.

I also used another display for the final version than the one I use on my breadboard and this display is extremely sensible to power fluctuation. If I have 4,8V then the display will dimm a lot and if the power is over 5,4V, the display will just go black (the contrast setting is very sensible).

I think the LDO is effectively the best solution in my case.

Grazie Luciano ;-)

[sougata]

Hi,

LDO would remove all headaches perhaps. By the time you get it consider this.

1. Use a simple 5.1V Zener shunt regulator for your PIC and LCD.
Both of them are not that current hungry. (Beware not the backlight)

2. Power the LEDs, Buzzer (and the LCD backlight if you are using) from the unregulated supply. (Batt +ve)

3. For lighting up LEDs or Buzzer invert your logic and drive the cathodes while the anodes are connected to Batt +ve through necessary limiting resistors.

[Luciano]

Hi,

Incorrect battery insertion is a common occurrence.

See this link:
FET Supplies Low-Voltage Reverse-Polarity Protection
http://www.elecdesign.com/Articles/I...ArticleID=9945

* * *

Also possible:

(Click to enlarge the picture).


Best regards,

Luciano

[flotulopex]

Sougata,

Why would it be better to "invert the logic" ?

Is it because the PIC can "sink" more current than "source" or is there any other reason?

[mister_e]

well i probably missed something but i don't see the advantage of it. Sure Sougata will explain why.

Usually you should initialize your hardware Somewhere at the Top of your code in a way that there's no current need. Let's say your LEDs are connected on PORTB between i/o and GND, you'll use....

PORTB=0
TRISB=0

For your Piezo... use a capacitor in serie. Directly it's just Killer. It may damage your PIC one day or another. 0.1 uF or higher is good.

ANY piezo sounder have there resonant frequency, if you select the right one, 0.1-0.47 uF is enough.

Do a seah for G-Spot on this forum... yeah no kidding!

The problem with command like SOUND, PWM, FREQOUT is that you never know if the PIN is set to HIGH or LOW at the end of the cycle. In case your PIEZO is between I/O and GND AND the SOUND command leave the I/o to HIGH, your Piezo will s u c k juice from your supply 'till you disable it the right way...

SOUND Piezo, ......
LOW PIEZO

[sougata]

Hi All,

If you are using a simple shunt regulator to power your PIC and the LCD (not the backlight) then while driving outputs high would require your shunt regulator to use a lower value resistor. That means to keep the zener under conduction with all the loads. When the loads are less the zener will act as the load. Keeping in mind that it is battery operated it results in higher quiscent current. I recommended using the LEDs,Buzzer and the backlight to be driven from the unregulated supply. This is only possible if you are using an inverted logic. And setting PORTB = 0 or PORTB = $FF takes the same amount of labor and code. Also the PIC and the LCD would be a fair enough steady load to dimension your resistor-zener shunt regulator for maximum supply efficiency. I hope I have made myself clear or am I missing something??? And using a cap with buzzer is always a good practice. To find out the root of the problem I suggested the experiment to find out exactly whats going wrong and where. Even if the problem is solved with software I would strongly recommend to use a cap as Steve mentioned. The buzzer is a killer.

[charudatt]

Hello Sougata,

I have a similar requirement, but without the LCD. Same 6V operation and a Piezo buzzer.

Its basically a code lock operating on 6V. A couple of flashing LED, Sounder (piezo) and a Solenoid.

I have a design before me which does not use even a series diode to drop the voltage, So I presume , the PIC can take that, till the Alkaline battery drops down to safer level on its own. Seeing this , I feel a series diode would be good.

I used a GP PNP transistor with zener at its base to regulate the voltage. Don't know if that's the right way.

Any help ?

[T.Jackson]

A shunt-based approach is possibly THE worst avenue you could take for a battery powered circuit. Zener diodes require at least 5mA of current at all time in order to maintain satisfactory regulation. As you can see, this is a highly undesirable attribute, particularly for batteries. Go with the suggested low dropout regulator - clearly your best option.